Eating Behaviors 14 (2013) 522–524
Contents lists available at ScienceDirect
Eating Behaviors
If it's good it must be bad: The indirect effect of temptation strength on self-control
through perceived unhealthiness
Floor M. Kroese ⁎, Catharine Evers, Denise T.D. de Ridder
Utrecht University, Department of Clinical and Health Psychology, P.O. Box 80.140, 3508 TC Utrecht, The Netherlands
a r t i c l e
i n f o
Article history:
Received 2 October 2012
Received in revised form 19 June 2013
Accepted 15 July 2013
Available online 21 July 2013
Keywords:
Food temptation
Temptation strength
Weight watching
Perceived healthiness
Self-control
a b s t r a c t
Previous research has shown that people tend to consume less from foods they consider more tempting.
However, the underlying mechanism for these counterintuitive findings is still unknown. The current paper is
the first to test the theoretically implied suggestion that the effect of food temptation strength on consumption
is indirect and can be explained through temptations' perceived unhealthiness. Two studies were conducted
among female students who were concerned about their weight to test the effect of food temptation strength
on perceived unhealthiness as well as the amount that was consumed of the products. Results showed that
temptation strength was associated with unhealthiness such that weak temptations were — unjustly — perceived
to be less unhealthy compared to strong temptations, while perceived unhealthiness was negatively related to
indulgence. As a consequence, people may consume more from weak than from strong temptations. It is concluded
that weak temptations tend to be underestimated and can be more challenging for successful self-regulation than
strong temptations.
© 2013 Elsevier Ltd. All rights reserved.
1. Introduction
The obesogenic food environment is often blamed for the rising
numbers of overweight in Western societies (French, Story, & Jeffery,
2001). The ubiquitous availability of unhealthy food temptations brings
along typical self-control dilemmas between immediate hedonic gratification and long-term health considerations. Although we may especially
fear the strong temptations that look most delicious, in this paper
we demonstrate that — and why — less attractive subtle temptations
may in fact be most dangerous.
Temptations, by definition, are in conflict with a (long-term) goal:
they are both attractive and forbidden (Kroese, Evers, & De Ridder,
2011). A traditional viewpoint in the literature is that food temptations
automatically trigger hedonic impulsive behavior and inhibit the
conflicting weight watching goal (Stroebe, Mensink, Aarts, Schut, &
Kruglanski, 2008). Counteractive control theory (Trope & Fishbach,
2000), however, proposes that rather than inhibiting goal pursuit,
(food) temptations activate the conflicting goal and trigger self-control
processes to protect it (Fishbach, Friedman, & Kruglanski, 2003).
In particular, such processes are suggested to display upon confrontation with strong temptations (i.e., those that present a large threat).
Indeed, prior studies have confirmed that strongly attractive food
temptations engender better self-control and consequently yield lower
consumption than weakly attractive food temptations (Kroese et al.,
2011). However, the underlying process for these intriguing findings
has not yet been demonstrated.
One suggested explanation is that weak temptations are perceived to
be less ‘threatening’. That is, weak food temptations would be believed to
be less bad in terms of weight watching goals (i.e., less unhealthy) and
consequently fail to trigger the alarm bells that signal the need for defensive self-regulatory action. Strong temptations, on the other hand, would
indirectly (through higher perceived threat) yield boosted self-control
performance.
The current aim is to provide empirical support for the implied indirect effect of temptation strength on self-control through temptations'
perceived threat, or unhealthiness. We expect that temptation strength
is positively related to perceived unhealthiness such that weak temptations are believed to be less unhealthy than strong temptations — even
when this is not objectively true. This is in line with the “unhealthy =
tasty intuition”, reported by Raghunathan, Naylor, and Hoyer (2006).
Additionally, we expect that perceived unhealthiness is negatively related
to consumption such that, indirectly, weak temptations yield higher
consumption compared to strong temptations.
In Study 1 we demonstrate that temptation strength is related to
perceived healthiness: weak temptations are considered less unhealthy
than strong temptations. In Study 2 we test the indirect effect of temptation strength on consumption through perceived unhealthiness.
2. Study 1
⁎ Corresponding author. Tel.: +31 30 253 1244.
E-mail address: [email protected] (F.M. Kroese).
1471-0153/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.eatbeh.2013.07.006
Study 1 was designed to demonstrate people's general tendency
to associate food temptation strength with unhealthiness.
F.M. Kroese et al. / Eating Behaviors 14 (2013) 522–524
523
2.1. Materials and method
3. Study 2
2.1.1. Participants
Fifty-nine female students participated. Data from one obese
participant (BMI N 30) were excluded from analyses, as obese people
are known to react differently to food stimuli (Stice, Spoor, Ng, & Zald,
2009). Participants in the final sample (N = 58) had a mean age of
21.5 years (SD = 1.6), a BMI of 21.8 kg/m2 (SD = 2.5), and were
moderately concerned about their weight (M = 4.18, SD = 1.36).
Study 2 aimed to demonstrate the indirect effect of temptation
strength on consumption through perceived unhealthiness. Temptation
strength was manipulated by exposing participants to either a large
bowl (strong temptation) or three small bowls (weak temptation)
of crisps (cf. Coelho do Vale, Pieters, & Zeelenberg, 2008). Perceived
unhealthiness was operationalized by assessing participants' estimates
of the caloric content of a handful of crisps. Furthermore, consumption
of the product was assessed. We hypothesized that weak temptations
yield lower caloric estimates, which in turn would lead to higher
consumption.
2.1.2. Procedure
The participants were presented with a short text describing a new
snack. Temptation strength was manipulated between-subjects by providing different taste ratings that were allegedly given by a test panel:
the test panel rated the tastiness of the snack at 6.0 (weak temptation)
or 8.5 (strong temptation) on a 10-point scale. After reading the text,
the participants rated the unhealthiness of the snack.
2.1.3. Material
2.1.3.1. Manipulation text.
A new snack is introduced to the market: a granola bar with pieces
of chocolate. […] The snack was recently presented to a test panel.
One question the panel was asked was how tasty they found the
new snack. The test panel rated the new snack on this question on
average at [6.0 — weak temptation condition/8.5 — strong temptation
condition] on a scale from 1 — not at all to 10 — extremely.
2.1.3.2. Estimated unhealthiness and attractiveness. The participants rated
“to what extent they thought the new snack would be healthy” [reverse
coded] on a scale from 1 (not at all) to 9 (very much). In addition, ratings
of the new snack's attractiveness served as a manipulation check for
temptation strength.
2.1.3.3. Control variables. Self-reported height and weight were assessed
to calculate BMI. Weight concern was assessed with one item (“To what
extent are you concerned about being slim”) on a scale from 1 (not at
all) to 7 (very much).
2.2. Results and discussion
2.2.1. Randomization and manipulation check
Two ANOVAs revealed no significant condition effect on BMI and
weight concern (p's N .33), indicating successful randomization.
An ANOVA testing the effect of condition on snack attractiveness
revealed a significant effect; F(1, 56) = 9.80, p = .003, ηp2 = .15.
Participants in the strong temptation condition (M = 5.50, SD = 1.69)
believed the snack was more attractive compared to participants in the
weak temptation condition (M = 4.03, SD = 1.87), indicating that our
manipulation was successful.
2.2.2. Estimated unhealthiness
An ANCOVA with condition as independent variable, estimated
unhealthiness as dependent variable, and weight concern as covariate, revealed a significant condition effect; F(1, 55) = 4.71, p = .03, ηp2 = .08.
Participants in the weak temptation condition rated the snack as less
unhealthy (M = 5.10, SD = 1.49) compared to participants in the
strong temptation condition (M = 5.82, SD = 1.47). The covariate
was significant; F(1, 55) = 4.52, p = .04, ηp2 = .08.
Study 1 thus confirmed our hypothesis that temptation strength
is associated with perceived unhealthiness: the mere suggestion that a
snack tastes good makes people believe it will be unhealthy.
3.1. Method
3.1.1. Participants
Forty-three female students participated. Data from 1 underweight
participant (BMI b 18) and 3 outliers (N3 SD) on either caloric estimates
or consumption were excluded from analyses. Participants in the
remaining sample (N = 39) had a mean age of 22.6 years (SD = 4.3),
a BMI of 21.9 (SD = 2.5), and were moderately concerned about their
weight (M = 3.9, SD = 2.0).
3.1.2. Procedure
The participants were exposed to either three small bowls of crisps
(weak temptation) or one large bowl of crisps (strong temptation).
The total amount of crisps was equal across conditions (100 g). The
study was presented as a marketing test and participants were asked to
taste and rate the crisps on several dimensions (appearance, crispiness,
etc.). Before consumption, participants were asked to estimate the number of calories in a handful of these crisps. Afterwards, consumption was
unobtrusively assessed.
3.1.3. Materials
3.1.3.1. Temptation strength. A pilot test (N = 29) confirmed that participants found the large bowl of crisps more tempting (M = 4.89,
SD = 1.11) than the three small bowls of crisps (M = 3.42, SD =
1.36); t(28) = 2.90, p = .007.
3.1.3.2. Bogus marketing test. The marketing test involved a number
of statements regarding the product. Participants were encouraged to
eat at least a little bit to be able to rate the taste of the product.
3.1.4. Main outcomes measures
3.1.4.1. Caloric estimates. Participants were asked to “estimate the number of calories in a handful of these crisps”. As a reference, the number
of calories in a cheese sandwich (150 Kcal) was given.
3.1.4.2. Consumption. The bowl(s) of crisps were unobtrusively weighed
before the participants entered the lab and after they had left. Postweight
was subtracted from preweight to calculate consumption.
3.1.4.3. Control variables. Age, height, current and ideal weight were
assessed by self-reports. Weight concern was assessed with 1 item:
“To what extent are you concerned about your weight”, on a scale
from 1 (not at all) to 7 (very much).
3.2. Results and discussion
3.2.1. Randomization check
One-way ANOVAs revealed no significant condition effects on age,
BMI, weight concern, or the amount of weight they wanted to lose
(p's N .35), indicating successful randomization.
524
F.M. Kroese et al. / Eating Behaviors 14 (2013) 522–524
3.2.2. Caloric estimates and consumption
Including weight concern as a covariate in all analyses, first caloric
estimates were regressed on temptation strength, showing that strong
temptations yielded higher caloric estimates compared to weak temptations (β = .39, p = .01). Furthermore, a regression of consumption on
caloric estimates showed that higher caloric estimates were associated
with lower consumption (β = –.33, p = .05). Bootstrapping analyses
confirmed that the indirect effect of temptation strength on consumption through caloric estimates was significant (95% CI: –4.44, –.15).
Study 2 thus showed that participants in the weak temptation
condition estimated lower numbers of calories in a handful of
crisps compared to participants in the strong temptation condition. Furthermore, lower caloric estimates were related to higher consumption. This finding supports our hypothesis that weak temptations
are unjustly believed to be less dangerous compared to strong temptations and consequently yield less effective self-control.
4. General discussion
The aim of this paper was to elucidate the processes underlying previous findings that strong temptations engender better self-control compared to weak temptations. Study 1 showed that temptation strength is
negatively related to perceived healthiness: a simple fact with regard to
the tastiness of a (hypothetical) snack affected its inferred healthiness.
Study 2 demonstrated that weak temptations, compared to strong temptations, yielded lower caloric estimates, which in turn led to higher
consumption.
Our findings provide empirical evidence for the theoretical implications put forward by counteractive control theory: To the extent that
temptations are a threat towards the weight watching goal defensive
self-control mechanisms will be activated. However, the threat (i.e., unhealthiness) of weak temptations tends to be underestimated and selfcontrol will be triggered to a lesser extent. Paradoxically, then, in some
cases weak temptations may have worse outcomes in terms of selfcontrol behavior compared to strong temptations.
The current studies are in line with previous findings suggesting that
weak temptations may be more challenging than strong temptations
(e.g., Coelho do Vale et al., 2008; Geyskens, Dewitte, Pandelaere, &
Warlop, 2008; Kroese et al., 2011). Furthermore, we add to accumulating
evidence suggesting that when trying to behave ‘good’ by taking
food that is not their most favorite temptation, such as small portions (Coelho do Vale et al., 2008) or food that seems rather healthy
(Chernev, 2011; Finkelstein & Fishbach, 2010), weight conscious people
often display ironic behavior (i.e., overconsume), making dietary failure
in fact more rather than less likely.
A seeming limitation is the selective population from which
our samples were drawn: young, weight conscious, female students.
However, our hypotheses only apply to people who are watching their
weight and trying to control their food intake. Female students are particularly likely to fulfill these criteria (Wardle, Haase, & Steptoe, 2006).
For unhealthy food to be a temptation, it needs to be attractive as well
as ‘forbidden’, or in conflict with a goal. Hence, for those who do not
have a weight watching goal, unhealthy food would not be an actual
temptation and we would not expect to find counteractive self-control
effects.
One avenue for future research would be to investigate ways to
counter the paradoxical effects of temptation strength on consumption.
A first plausible direction could be to make people explicitly aware of
the actual healthiness or caloric content of foods, such that estimated
healthiness can no longer play a role.
To conclude, the current research provides insight into the underlying
processes explaining the effects of temptation strength on self-control. It
was demonstrated that people assume a positive relationship between
temptation strength and unhealthiness, such that weak temptations
are — sometimes unjustly — perceived as less unhealthy. Consequently,
dieters paradoxically may fail, not because the food ‘was just too good’
but because it was only ‘just good’.
Role of funding sources
Funding for this study was provided by a Toptalent grant to the first author from
The Netherlands Organization for Scientific Research (NWO). NWO had no role in the
study design, collection, analysis or interpretation of the data, writing the manuscript,
or the decision to submit the paper for publication.
Contributors
All three authors were involved in the design of the study. FK conducted the statistical
analyses and wrote the first draft of the manuscript. All authors contributed to and approved
the final manuscript.
Conflict of interest
The authors have no conflict of interest to declare.
References
Chernev, A. (2011). The dieter's paradox. Journal of Consumer Psychology, 21, 178–183.
Coelho do Vale, R., Pieters, R., & Zeelenberg, M. (2008). Flying under the radar: Perverse
package size effects on consumption self-regulation. Journal of Consumer Research,
35, 380–390.
Finkelstein, F. R., & Fishbach, A. (2010). When healthy food makes you hungry.
Journal of Consumer Research, 37, 357–367.
Fishbach, A., Friedman, R., & Kruglanski, A. (2003). Leading us not unto temptation:
Momentary allurements elicit overriding goal activation. Journal of Personality and
Social Psychology, 84, 296–309.
French, S. A., Story, M., & Jeffery, R. W. (2001). Environmental influences on eating and
physical activity. Annual Review of Public Health, 22, 309–335.
Geyskens, K., Dewitte, S., Pandelaere, M., & Warlop, L. (2008). Tempt me just a little
bit more: The effect of prior food temptation actionability on goal activation and
consumption. Journal of Consumer Research, 35, 600–610.
Kroese, F. M., Evers, C., & De Ridder, D. T. D. (2011). Tricky treats: Paradoxical effects of
temptation strength on self-regulation processes. European Journal of Social Psychology,
41, 281–288.
Raghunathan, R., Naylor, R. W., & Hoyer, W. D. (2006). The unhealthy = tasty intuition and its effects on taste inferences, enjoyment, and choice of food products. Journal of Marketing, 70, 170–184.
Stice, E., Spoor, S., Ng, J., & Zald, D. H. (2009). Relation of obesity to consummatory and
anticipatory food reward. Physiology & Behavior, 97, 551–560.
Stroebe, W., Mensink, W., Aarts, H., Schut, H., & Kruglanski, A. W. (2008). Why dieters fail:
Testing the goal conflict model of eating. Journal of Experimental Social Psychology, 44,
26–36.
Trope, Y., & Fishbach, A. (2000). Counteractive self-control in overcoming temptation.
Journal of Personality and Social Psychology, 79, 493–506.
Wardle, J., Haase, A.M., & Steptoe, A. (2006). Body image and weight control in young adults:
International comparisons in university students from 22 countries. International Journal
of Obesity, 30, 644–651.